Vapourware Codexian Game Development Thread

[The Avatar]

Do they actually do the code signing/notarization for you so Gatekeeper doesn't complain? Or do you have to pay Apple $99 a year for that privilege?

Unity is the same way, you just switch platforms and build.

 

[zwanzig_zwoelf]

Do they actually do the code signing/notarization for you so Gatekeeper doesn't complain? Or do you have to pay Apple $99 a year for that privilege?

Unity is the same way, you just switch platforms and build.

IIRC signing became mandatory around late 2018, but I'm not sure if you're still blocked from running an application if you try to do it w/ admin rights.

Also it's not that simple, MacOS is more restrictive with what you can do compared to Windows. Yes, you can build, but when it comes to running the damn thing, you'll probably face restrictions, platform-specific bugs, and other 'fun' stuff.

Then again, same or similar things can be said about any other platform.

 

[Bad Sector]

I spent some time with that Quake 1 map editor i started last year (and got bored after a few days) to fix some bugs (can now work in Win98 on a Pentium 3 retro PC i have with a Voodoo 3) and implement map file saving (i already had implemented map loading before). Now it has enough functionality to make simple working maps for Quake, though it still is missing some essential functionality like texture alignment and being able to clip brushes (for making ramps, non-square corners, etc - this is why everything is boxy in the screenshot). Might do that next year or whatever :-P.

 

[RobotSquirrel]

Doing loypoly collars is surprisingly difficult. The problem is that the bind seems to want to pull the collar into the clothing and the geo is all kinds of wack, a work around was to add more geo so that the collar has a lip of sorts.
My head hurts lol. I mean the character is basically done excluding textures its just this damn collar! arrr! lol.

 

[Bad Sector]

In a slightly less retro tune, i did some contact hardening shadow filtering in Little Immersive Engine:

It is kinda slow though, at least with 8xMSAA. It is a bit faster without antialiasing but looks more grainy:

Not sure if i'll keep it though. I'll try to optimize it a bit at some point later, but if i can't get it fast enough (ie. being only a little bit slower than without the filter) i'll probably remove it. Looks nice though.

 

[infidel]

https://www.youtube.com/c/Kingsblood/videos

What's this indie game dev obsession with video dev logs? This guy is up to his 51st dev log, average video length is like 15 minutes. Must take hours for each one to write a script and present and edit... and he only gets like 100 views per video. Is it worth it? It looks like a huge waste of time.

You overestimate the efforts that went into those. The recent ones are more like 5-7 minutes and feature some rambling and a recording of his PC screen. The ones that are longer have even more rambling and him mostly staring at Unity. If he got the grip of it, maybe an hour or two of recording/editing per video? And these are spread over 12 months, looks like he's aiming at one video per week. It's worth it in the sense of learning all the basics of video production and quality rambling while secretly hoping to get at least a small-sized ball rolling :D. Whether that transforms into something tangible, only time will tell. But it has its fun. I've recently bought Vegas and currently experimenting with various music-related videos and it is a fun challenge - making something that both looks good and engages the viewer. Of course, I can't claim any success whatsoever

 

[infidel]

Easy, right? Set up a VPN, share my files with the Mac, run Unreal Build Tool, and I've got myself a Mac build. Except Egypt blocks VPNs.

Okay, I have to ask. Can't you spin up your own personal VPN? Or does Egypt block OpenVPN too?

 

[Nathaniel3W]

Easy, right? Set up a VPN, share my files with the Mac, run Unreal Build Tool, and I've got myself a Mac build. Except Egypt blocks VPNs.

Okay, I have to ask. Can't you spin up your own personal VPN? Or does Egypt block OpenVPN too?

That's exactly what I was using. I don't want to pay money to use someone else's VPN, and as far as I know, those are only useful for browsing the internet anyway. I used to run OpenVPN on an Amazon EC2 instance, and I would connect to my VPN in order to access my bank, my insurance, and other stuff that's touchy about logging in from outside the US. But eventually my VPN stopped working. Sometimes I could get it to work again by shutting it down and restarting it on a different IP address. I tried getting it working again later for this project, and I couldn't get the VPN to work at all. I can connect to my cloud computers via SSH and RDP, and I can use them as my proxy server. But I can't connect to them as a VPN.

 

[Nathaniel3W]

In a slightly less retro tune, i did some contact hardening shadow filtering in Little Immersive Engine:...

Cool. So it makes the shadow fuzzier the farther away it is from the object that cast it? I don't think I've noticed any other engines doing that. Are you eyeballing it until it looks right, or are you following a real physics textbook?

 

[Bad Sector]

Cool. So it makes the shadow fuzzier the farther away it is from the object that cast it? I don't think I've noticed any other engines doing that. Are you eyeballing it until it looks right, or are you following a real physics textbook?

Yeah that is what it does. It is me eyeballing it - it is just a linear scale for how much to spread the extra sampling positions based on the distance from the shadow blocking surface. The spread is also capped so that it doesn't go too big as it can introduce glitches when two shadows of very different spreads intersect (there are ways to fix that but they slow down things even more).

The technique isn't anything new, i actually first implemented it back in 2007 or so when i worked on Theseis - that was me eyeballing it again, but others tried similar approaches - i think Hellgate London had (or planned to have, i'm not sure if it made it in the final game) something similar (though i wasn't aware of it at the time) - and i did another implementation around 2012. But in recent years it has become quite common, AFAIK both Unity (via HDRP) and UE4 support them and it was one of the features in Nvidia's Gameworks (specifically the ShadowWorks package).

Of the recent games i know using them is Cyberpunk 2077 (it was actually prototyped in Witcher 3 but didn't make it to the final game). Sadly i can't find an easy screenshot for the shadows since all shots i find seem to be about raytraced shadows... but i CAN find a screenshots when the algorithm fails (the glitch i mentioned previously) :-P

Spoiler

Notice how the shadow "breaks" between the shadow for the hair and the shadow from the cards (visible especially on the right card on the table from our perspective). It is more obvious in motion.

It is a neat effect but it is quite slow and somewhat glitchy. During Theseis development the idea was to only use it very sparingly (it'd be a "high end" PC-only feature) for dramatic lighting (e.g. something like a large light casting on a room through a door or down a hall so you can see a character walking away from the light and see the shadow becoming harder as he approaches the wall) and only one such shadow would be visible - others would be regular shadows and for some cases even avoiding real shadows at all and using projected shadows like Half-Life 2 did.

Computers are faster now though and i'd like to avoid special cases in the renderer since i want to keep the engine (relatively) simple. After all the entire purpose of the engine (aside personal satisfaction :-P) is to make a very easy (for me) to use engine by abusing modern hardware to do mid-2000s graphics :-P.

EDIT: so i optimized the shadows a bit, now they're about 42% faster:

Spoiler

This used to be ~45fps. There are with barely any visual differences, meaning they're still grainy, but whatever - i'll see about that later. Or leave it to MSAA :-P.

Spoiler

Or implement TAA

 

[infidel]

I can connect to my cloud computers via SSH and RDP, and I can use them as my proxy server. But I can't connect to them as a VPN.

Well, SSH tunnel is mostly the same, I guess.

 

[Bad Sector]

Ok, i think i'll keep the soft shadows though they're only going to be used for high settings. I added proper configuration and variations in the lighting code so that there are several quality and filtering settings.

With that in mind i also tried to run the engine again on my old Athlon 64 X2 PC with an AGP X1950 Pro - this is almost 15yo hardware so i guess if it runs there, it'll run everywhere (even if in low settings) :-P. And, after the driver exposed a bug (see, it is a good idea to have a bunch of PCs around with different GPUs etc to try stuff) i actually made it run, though...

...the performance wasn't that great. Fortunately this is largely because of the large number of objects (~75374 meshes) and lights (~2028) over the 16900 square meter scene and even though only a single room is visible (the room is repeated to cover the entire scene area) the renderer does an "is this object/light visible?" check for every single one of them, so the rendering is actually CPU limited, not GPU limited (for this scene at least anyway). I wrote "fortunately" because the current occlusion culling approach is temporary and i do plan on improving it, but i need to work on the tool side first because it'd require some world editor support - and there isn't really a world editor there yet :-P.

But as a quick test i removed most of the smaller meshes and i got a more agreeable performance (there are still a few thousand meshes in the scene so an improved occlusion culling system it should still improve things):

And just for the sake of trying it out, i disabled shadows for a potential "low settings" mode and got 60+ fps:

So yeah, if nothing else the engine can run in old PCs - at least in "ugly mode" :-P

 

[Nathaniel3W]

I'm working on some HUGE changes to my game. These changes are going to be so big I think I'll have to include them as optional, like in a Steam Workshop item or a free DLC or something. I'm still trying to decide on what exactly I want to do. I'll release more info as development continues.

For now, check out the new menu system I'm working on. My current (old) menu system is a complete mess. When I wrote it, I really didn't know what I was doing. Basically, in the old system, I had a bunch of HUGE Flash swf files, with all the code written directly onto the timeline. Any place where you might want to access the same information (like going to options from the main menu, and options from the pause menu; or going to the character sheet from the pause menu, and character sheet from the equipment vendor) I just duplicated that part of the menu. And if I ever had to change it, I had to change multiple copies in different menus.

Also, the old system doesn't work well with controllers. I made the controllers simulate mouse input, which is definitely not the ideal approach.

This menu you see here is completely new, reusing none of the code from the old menu. You can switch from mouse to controller instantly and seamlessly: if you press a controller button, it takes effect, and the mouse disappears; likewise if you had been using the controller and then you move the mouse, the mouse cursor appears and you can use it instantly. Also, the new menu system works with a system of layers. Mouse on top, HUD on the bottom, and any possible menu screen will go somewhere in between. Because it works on a system of layers, I can open any menu from any other menu, and it will just display as a new layer. And when I close the top layer, it just takes me back to the previous layer. And I can customize it so the layer either blocks layers below, or it can not block lower layers, and you can interact with multiple layers at once.

And with how modular it is, I've already started reusing code. The menu bar on the left side of the screen uses the exact same code as a game-exit confirmation window displayed in the middle of the screen. To make a different menu, I only have to provide different button names and hook them up to different functions. The side menu can be the pause menu, or a shop, or an inn, or a tavern, or anything else, with different numbers of buttons created as needed. A menu in the middle of the screen that has a menu name and any number of buttons can just reskin the side menu.

This is far from the most exciting part of game development. But still, it feels good to go back and do something the right way, and to know why this is the right way.

 

[Bad Sector]

I had a bunch of HUGE Flash swf files

Out of curiosity, how did you make the SWF files and integrate them in the game? AFAIK both Flash and Scaleform are long dead (in that Adobe hasn't made any new versions for a long time now and many engines seem to remove any support).

I made the controllers simulate mouse input, which is definitely not the ideal approach.

Don't worry, Cyberpunk 2077 does the same thing, so you can at least claim AAA quality standards :-P

 

[Tavernking]

On itch.io you can do 'reverse-sales', where you temporarily increase the price of the game. Makes you think...

 

[Nathaniel3W]

Out of curiosity, how did you make the SWF files and integrate them in the game? AFAIK both Flash and Scaleform are long dead (in that Adobe hasn't made any new versions for a long time now and many engines seem to remove any support).

Because I'm working in Unreal Engine 3, which I believe was last updated in 2015, and which has an integrated Flash player. If Windows ever refuses to work with any embedded Flash player, then that would destroy virtually every game made from 2008-2011. (Huh. I just looked through this Wikipedia list. Borderlands 3, made in 2019, used Scaleform.) The version of Animate that I'm using still supports ActionScript 3, but ActionScript 2 is gone, which breaks all of the tutorials and example files that come with UDK. Adobe will still let you download old versions of Flash Pro if you have a Creative Cloud subscription. It's kind of a pain, but they'll let you do it.

You're pretty much correct though that Flash and Scaleform are long dead and unsupported. It's pretty hard for me to find help and documentation examples [edit: the AS3 documentation is actually really good], but the Epic Forums and Stack Overflow have always seemed to pull through for me when it matters.

Don't worry, Cyberpunk 2077 does the same thing, so you can at least claim AAA quality standards :-P

Ha! Thanks. Seriously though, that is really surprising. PC may have been the largest single platform for Witcher 3, but most people played on consoles. So counting 100% of console players using controllers, plus some percentage of PC players using controllers, CDPR's biggest success was played mostly on controllers. You'd think they would design their games with controllers first in mind.

 

[zwanzig_zwoelf]

DGJ is getting ready to wrap up within a month or two, and I'm already thinking about the next project. This time I'll probably do a traditional Wiz-style dungeon crawler and hire some outside help to handle the enemy/event graphics.

One thing I'm not 100% sure about yet is whether to keep the simple wireframes (Wiz 1-5) or use a more elaborate (DG-style) approach with environmental detail. I guess I'll have to experiment...

 

[Tavernking]

One thing I'm not 100% sure about yet is whether to keep the simple wireframes (Wiz 1-5) or use a more elaborate (DG-style) approach with environmental detail. I guess I'll have to experiment...

Environmental detail please. For people who never played dungeon crawlers growing up everytime you post it just looks like you're showing us a bunch of rectangles

 

[Tavernking]

Unrelated, here is my latest project

 

[ProphetSword]

I know I tried something similar when I started working on my game...

I was told pretty emphatically by people on this forum that nobody would play a game with wireframe walls in this day and age. Doesn't seem like it is a popular type of game anymore.

 

[zwanzig_zwoelf]

Environmental detail please. For people who never played dungeon crawlers growing up everytime you post it just looks like you're showing us a bunch of rectangles

Acknowledged this user's agenda.

I was told pretty emphatically by people on this forum that nobody would play a game with wireframe walls in this day and age. Doesn't seem like it is a popular type of game anymore.

Same people told you that games without romance and dialogue aren't true RPGs?

Just kidding. It's an uncommon style, but it's a good way to stand out among 9000+ samey pixel games.

I think the perception that 'nobody would play' this kind of games comes from two things: a very small number of games released in this style over the past ~10 years and lack of very popular entries.

 

[Lexx]

I made some Fo2 mini-mods yesterday and the day before: Lightweight carry weight indicator and Fo2 timer notes.

The first one works with any Fo2 mod, as long as Sfall is not ancient, and the second obviously only works for Fo2 itself (vanilla/UP/RP). Both are more or less just-for-fun project, becaues I wanted to do something small and quick. I think they turned out pretty nice, though I doubt people will use them much. Who knows. :>

 

[ProphetSword]

Same people told you that games without romance and dialogue aren't true RPGs?

People have a lot of opinions. Great thing about developing games is that you learn after a while that some comments are valuable and some should be ignored.

Just kidding. It's an uncommon style, but it's a good way to stand out among 9000+ samey pixel games.

The rule I've always lived by is this: Make the game you would want to play. If you want to play it, there are others who will want to play it as well. And they will appreciate the passion you put into it. Anyone who doesn't get it isn't your target audience anyway.

I think the perception that 'nobody would play' this kind of games comes from two things: a very small number of games released in this style over the past ~10 years and lack of very popular entries.

You know, if anyone can pull it off, my money is on you. I hope you prove everyone wrong.

 

[rusty_shackleford]

I was told pretty emphatically by people on this forum that nobody would play a game with wireframe walls in this day and age. Doesn't seem like it is a popular type of game anymore.

Making a game that will be popular is a fool's errand. For every minecraft, rimworld, etc., there are over a thousand completely unknown games, and tens of thousand(or probably more) games that never see the light of day.
Being an indie dev won't make you rich. If you're a programmer, you can go make lucrative money in a different field. If you're an artist, guess you're stuck drawing porn.

The biggest priority is working on a project you enjoy working on. If your goal is to make money... making video games is not the way to do it.

 

[The Avatar]

I disagree. Plenty of indie games have been successful enough to have made their creators fairly wealthy. Just not indie games with wireframe walls.